A Machine Learning Method for Predicting Loan Approval by Comparing the Random Forest and Decision Tree Algorithms.

Abstract

Aim: The objective of this work is to determine an approach in machine learning for loan approval prediction by comparing Random Forest algorithms with Decision Trees. To achieve accuracy novel random forest classifiers are used. Materials and Methods: Loan prediction datasets from the kaggle library are used to test accuracy and loss.The total sample size is 20. The two groups considered were Random Forest (N=10) and Decision tree (N=10). The computation is performed using G-power as 80%. Results: While the random forest method has a precision of 79.4490% and loss is 21.0310%, a method that looks superior to the traditional decision tree of 67.2860% and loss is 32.7140% respectively. Finally, it seems that the Random Forest method outperforms the Decision tree. RF and DT, The independent sample T-test result of p=0.33 (p>0.05) shows a statistically significant agreement between the two most extensively used machine learning techniques shows that two groups are statistically insignificant with confidence level of 95%. Conclusion:Random Forests seem to be more accurate in predicting loan acceptance than Decision Trees.